74 SCIENCE PROGRESS 



It will be observed that the initial loss of water from the 

 crimson Salicornia and from Suceda is higher than that from 

 Mercurialis and the green Salicornia ; this is due to the fact 

 that the weighings of the two former were begun later in the 

 day than those of the two latter plants. 



The experiments show that the amount of water transpired 

 from the halophytes varied pretty considerably, but in all 

 cases it was greater than for a typical mesophyte ; that the 

 maximum rate of transpiration per hour was attained by 

 Suaida ; and that the transpiration of the crimson Salicornia 

 was not higher than that of the green variety, a result which 

 is contrary to those of Chandler's observations and which may 

 be due to their being taken from a different locality. 



If the succulence of a plant be taken as more or less equiva- 

 lent to the water content per ioo sq. cm. of surface, then 

 Suazda is the most succulent of the plants employed in the 

 above experiment. This is shown by the following figures 

 (Table V) : 



Fresh weight 



Dry weight at ioo'C. . 



Water content per cent. 



Surface area . . . 45 "14 sq. cm. 24-43 sq. cm. I7"44sq. cm. 4038 sq. cm. 



Water content per iool . „ ,.„ ,. 



cm ) 95 gm. 6-5 gm. 6 1 gm. 10 gm. 



It therefore follows that succulence is not necessarily bound 

 up with a reduced transpiration. 



Further experiments on these lines were carried out in 

 departmental laboratories where more precautions could be 

 taken than in the laboratory at Erquy. 



Shoots of Salicornia annua and leaves of Sedum spurium 

 and of Vicia cracca were selected, and the cut ends sealed with 

 wax. They were suspended, by means of a loop of cotton, 

 in air saturated with water vapour until their turn came to be 

 weighed. After their weights had been noted they were sus- 

 pended in the relatively still air of the laboratory, all under 

 the same conditions, together with a weighed dish of water. 

 The dish and the plants were periodically weighed. The fol- 

 lowing table (VI) summarises the results, in which the transpira- 



